Last Comment Bug 689118 - Firefox- Crash Report [@ js::AutoCompartment::AutoCompartment ]
: Firefox- Crash Report [@ js::AutoCompartment::AutoCompartment ]
Status: VERIFIED FIXED
[firebug-p1]
: crash, topcrash
Product: Core
Classification: Components
Component: JavaScript Engine (show other bugs)
: Trunk
: x86 All
: -- critical with 12 votes (vote)
: mozilla12
Assigned To: Blake Kaplan (:mrbkap)
:
Mentors:
https://crash-stats.mozilla.com/repor...
: 701130 713251 716426 716839 716849 718378 720348 (view as bug list)
Depends on: 712289
Blocks: 628758
  Show dependency treegraph
 
Reported: 2011-09-26 02:25 PDT by Yann Brelière
Modified: 2012-02-06 11:45 PST (History)
31 users (show)
See Also:
Crash Signature:
QA Whiteboard:
Iteration: ---
Points: ---
Has Regression Range: ---
Has STR: ---
affected
+
verified
-
verified
-
verified


Attachments
GDB output (2.75 KB, text/plain)
2011-10-20 11:25 PDT, Brent Garber
no flags Details
Correct GDB output (67.93 KB, text/plain)
2011-10-20 11:37 PDT, Brent Garber
no flags Details
Proposed fix (1.93 KB, patch)
2012-01-03 07:44 PST, Blake Kaplan (:mrbkap)
luke: review+
akeybl: approval‑mozilla‑aurora+
akeybl: approval‑mozilla‑beta+
mrbkap: checkin+
Details | Diff | Splinter Review
A script that always triggers a crash for me (102 bytes, text/html)
2012-01-16 01:15 PST, Ariel Shkedi
no flags Details

Description Yann Brelière 2011-09-26 02:25:19 PDT
User Agent: Mozilla/5.0 (X11; Linux i686; rv:9.0a1) Gecko/20110925 Firefox/9.0a1
Build ID: 20110925172321

Steps to reproduce:

On my session, in Firefox Nightly, every time I open https://www.gandi.net (or the dev or preprod website, or another page on the site), it crashes:


Actual results:

https://crash-stats.mozilla.com/report/index/bp-9c67f267-0a39-495f-8549-4e2282110926
https://crash-stats.mozilla.com/report/index/bp-a6e13aa8-7105-4d93-b51d-f1f272110926


Expected results:

In the nightly of the 21 or 22 of september, it worked fine.
Comment 1 Yann Brelière 2011-09-26 02:28:45 PDT
The crash doesn't happen in another Nightly session or in private browsing mode. I have removed all my cookies for gandi.net and the crash still happens.
Comment 2 Yann Brelière 2011-09-26 02:46:48 PDT
It's a Firebug bug in fact: http://code.google.com/p/fbug/issues/detail?id=4861
If I deactivate Firebug, it doesn't crash anymore.
Comment 3 Robert Kaiser 2011-09-28 13:19:28 PDT
No idea what "Firefox-Trunk" is, but it also happens enough in Firefox Nightlies.
Comment 4 Yann Brelière 2011-09-28 13:32:00 PDT
Yes, it is Firefox Nightly. I guess it is called like that because the package in Ubuntu is firefox-trunk (this allows running firefox alongside): https://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa.
Thanks for confirming.
Comment 5 Jan Honza Odvarko [:Honza] 2011-09-29 05:36:58 PDT
I am marking as Firebug p1, since it's blocking Firebug on nightlies.

Could this problem be related to Bug 689101?
It looks like it appeared in the same regression range.

Honza
Comment 6 :aceman 2011-09-30 05:41:48 PDT
I guess comment 3 confirms the crash is valid.
Comment 7 Yann Brelière 2011-10-17 03:12:12 PDT
Great! I don't have the crash anymore, and Firebug works. I believe this is thanks to Bug 689101, so I don't know if this bug should be marked as a duplicate or as depending on Bug 689101.
Comment 8 Jan Honza Odvarko [:Honza] 2011-10-17 03:23:21 PDT
(In reply to Yann Brelière from comment #7)
> Great! I don't have the crash anymore, and Firebug works. I believe this is
> thanks to Bug 689101, so I don't know if this bug should be marked as a
> duplicate or as depending on Bug 689101.
Thanks for the update!

Honza
Comment 9 Steve Fink [:sfink] [:s:] 2011-10-17 09:54:58 PDT
(In reply to Yann Brelière from comment #7)
> Great! I don't have the crash anymore, and Firebug works. I believe this is
> thanks to Bug 689101, so I don't know if this bug should be marked as a
> duplicate or as depending on Bug 689101.

I'll mark it as a duplicate. I'm glad it fixed this for you!

*** This bug has been marked as a duplicate of bug 689101 ***
Comment 10 Brent Garber 2011-10-20 11:23:23 PDT
I'm not sure this is a duplicate of the other ticket, I'm still receiving crashes related to firefox nightlies and firebug javascript breakpoints.

I'll attach gdb output.
Comment 11 Brent Garber 2011-10-20 11:25:51 PDT
Created attachment 568449 [details]
GDB output

Mozilla/5.0 (Ubuntu; X11; Linux i686; rv:10.0a1) Gecko/20111017 Firefox/10.0a1

firefox-trunk-10.0~a1~hg20111016r78825

https://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa
Comment 12 Brent Garber 2011-10-20 11:37:21 PDT
Created attachment 568453 [details]
Correct GDB output

Of course I accidently overwrote the GDB output when it crashed in AutoCompartment, but still settingt breakpoints causes various crashes. I'll try to gather more, but most seem related to cyclecollector
Comment 13 Yann Brelière 2011-10-28 02:47:09 PDT
I can confirm that this bug was in fact, not resolved.
I was thinking it was, because I could open Firebug and activate the console and the script panel without crashing.
But next time I start Nightly and open firebug on the tab where I activated the console, it crashes.
Comment 14 Jan Honza Odvarko [:Honza] 2011-10-28 02:52:36 PDT
Yann, can you please attach your crash report?

Honza
Comment 15 Yann Brelière 2011-10-28 03:00:44 PDT
Do you know how I could do that? Firefox on ubuntu has removed crash-reporter (see bug 694950).
Comment 19 Nigel Babu [:nigelb] 2011-11-24 23:17:07 PST
I just had a similar crash.  Report - https://crash-stats.mozilla.com/report/index/bp-9ab3f2fc-d160-4843-b6fb-8dff02111124
Comment 20 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-03 07:09:54 PST
This is 100% reproducible for me - opening a specific page immediately takes down Firefox 9.0.1 for me, if Firebug is enabled.

https://crash-stats.mozilla.com/report/index/bp-6d51c202-96a6-4c0b-b3bf-8ec292120103
https://crash-stats.mozilla.com/report/index/bp-388cd0cf-2c20-4195-85a3-544352120103
https://crash-stats.mozilla.com/report/index/bp-4c18aef9-dce1-48db-8029-c158a2120103
Comment 21 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-03 07:13:57 PST
*** Bug 713251 has been marked as a duplicate of this bug. ***
Comment 22 :aceman 2012-01-03 07:24:49 PST
Can you give the URL for that specific page?
Comment 23 Blake Kaplan (:mrbkap) 2012-01-03 07:41:46 PST
I bet this has to do with the value debugged being null.
Comment 24 Blake Kaplan (:mrbkap) 2012-01-03 07:44:49 PST
Created attachment 585410 [details] [diff] [review]
Proposed fix

I snuck in a warning fix since jsd is C and includes jsprvtd.h (via jsdbgapi.h, who knew). Looking further through this file shows that everywhere else null-checks the object if it uses JSVAL_IS_OBJECT.

I eagerly await the day that we can use JS::Value in jsd2.
Comment 25 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-03 07:52:52 PST
(In reply to :aceman from comment #22)
> Can you give the URL for that specific page?

It's internal, sadly. I can put it up with the necessary licensing/copyright disclaimers that the responsible people require sometime tomorrow (it's end of business here in NL right now). Sorry!

My work machine is building 9.0.1 with  attachment 585410 [details] [diff] [review] right now. I'll test that ASAP tomorrow morning.
Comment 26 Luke Wagner [:luke] 2012-01-03 08:30:22 PST
Comment on attachment 585410 [details] [diff] [review]
Proposed fix

Word.
Comment 28 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-04 08:50:26 PST
Ah, I forgot to mention this here. I tested builds with/without the fix, and this fixes the problem for me.

Blake: what would be necessary for this to make Fx 10/11? It's a pretty annoying bug for people hitting it...
Comment 29 Ed Morley [:emorley] 2012-01-04 17:29:19 PST
https://hg.mozilla.org/mozilla-central/rev/5025534b9d88
Comment 30 Scoobidiver (away) 2012-01-05 14:44:56 PST
*** Bug 701130 has been marked as a duplicate of this bug. ***
Comment 31 Scoobidiver (away) 2012-01-05 14:50:32 PST
With combined signatures, it's #1 top crasher in 9.0.1 on Mac OS X with 25% of all Mac crashes, so I think it should be fixed in Aurora and Beta.
Comment 32 Leho Kraav (:macmaN @lkraav) 2012-01-08 04:56:40 PST
is bug 674568 related to this? having firebug enabled for my development process crashes firefox 9.0.1 constantly after testing form submits. is there anything i can do to hold me over until this patch lands?
Comment 33 Matthias Versen [:Matti] 2012-01-08 18:01:02 PST
*** Bug 716426 has been marked as a duplicate of this bug. ***
Comment 34 Anael Mobilia 2012-01-08 18:42:50 PST
Should it really wait to FF12 version to fix this bug ?

Many users appears to have this problem as reported on the crash stats :
https://crash-stats.mozilla.com/report/index/0b0851c4-5e1d-4ac4-92f2-6685c2120106 -> comments

This bug can be reached by the 3 000 000 people who use FireBug add-on each day.
https://addons.mozilla.org/fr/firefox/addon/firebug/statistics/?last=365

Isn't it a customer satisfaction rate ?

Regards,
Anael
Comment 35 Steve Fink [:sfink] [:s:] 2012-01-09 10:50:52 PST
Comment on attachment 585410 [details] [diff] [review]
Proposed fix

[Approval Request Comment]
Regression caused by (bug #): unknown

User impact if declined: Firebug (or another JSD client) crashes the browser in some circumstances

Testing completed (on m-c, etc.): commenters have tried it out in m-c

Risk to taking this patch (and alternatives if risky): minimal. It only matters when jsdval->val is null, and without the patch it's an immediate crash in that case. The new behavior is to pick a fallback object as the scope object, and it's the same fallback that is already used for numeric or other non-object values.
Comment 36 Blake Kaplan (:mrbkap) 2012-01-09 10:57:06 PST
Comment on attachment 585410 [details] [diff] [review]
Proposed fix

[Approval Request Comment]
Regression caused by (bug #): bug 628758
User impact if declined: Firebug users will crash when debugging null values.
Testing completed (on m-c, etc.): Gijs has tested to make sure that this fixes the crash.
Risk to taking this patch (and alternatives if risky): This patch is as un-risky as it gets. It adds a null check and avoids crashing if the given value is null.

Note that the comment-fix part of this patch (js/src/jsprvtd.h) doesn't apply to aurora or beta and therefore won't be checked in. The comments being fixed don't exist on that branch.
Comment 37 Alex Keybl [:akeybl] 2012-01-09 11:13:41 PST
Comment on attachment 585410 [details] [diff] [review]
Proposed fix

[Triage Comment]
Approving this top crasher with minimal risk (thanks for the background Steve and Blake) for both Aurora and Beta. Please land today to make the 4th beta. If this misses tomorrow's beta, we may have to reevaluate landing it for FF10.
Comment 38 Jan Honza Odvarko [:Honza] 2012-01-09 11:50:54 PST
(In reply to Blake Kaplan (:mrbkap) from comment #36)
> Comment on attachment 585410 [details] [diff] [review]
> Proposed fix
> 
> [Approval Request Comment]
> Regression caused by (bug #): bug 628758
> User impact if declined: Firebug users will crash when debugging null values.
> Testing completed (on m-c, etc.): Gijs has tested to make sure that this
> fixes the crash.
> Risk to taking this patch (and alternatives if risky): This patch is as
> un-risky as it gets. It adds a null check and avoids crashing if the given
> value is null.
> 
> Note that the comment-fix part of this patch (js/src/jsprvtd.h) doesn't
> apply to aurora or beta and therefore won't be checked in. The comments
> being fixed don't exist on that branch.
Is there any workaround for Firebug (on Firebug side) to avoid the crash?
Honza
Comment 39 Blake Kaplan (:mrbkap) 2012-01-09 12:03:15 PST
I don't think so. The bug is pretty low-level in JSD. The closest thing I can think of is to not debug objects that have null values as properties.
Comment 41 Anael Mobilia 2012-01-09 12:08:44 PST
At first view : not for an "[@ js::AutoCompartment::AutoCompartment ]" bug.

But is it really a Firebug error ?
Really low level problem (which should be prevented by JS daemon : a typo on an extension should not result to a brutal crash of the whole Firefox !

Regards,
Anael
Comment 42 Jan Honza Odvarko [:Honza] 2012-01-09 22:32:02 PST
(In reply to Anael from comment #41)
> But is it really a Firebug error ?
> Really low level problem (which should be prevented by JS daemon : a typo on
> an extension should not result to a brutal crash of the whole Firefox !
Sure, agree.

I just wanted to do something for users running Firefox without the patch.

Honza
Comment 43 Scoobidiver (away) 2012-01-10 02:28:46 PST
*** Bug 716839 has been marked as a duplicate of this bug. ***
Comment 44 Jan Honza Odvarko [:Honza] 2012-01-10 06:25:49 PST
What is currently the target Firefox version for the patch?
I am getting quite a lot of questions about when this will be fixed.
Honza
Comment 45 :aceman 2012-01-10 06:28:56 PST
It looks like it got into Firefox 10, 11 and 12 (per comment 40 and comment 29).
Comment 46 Scoobidiver (away) 2012-01-10 06:34:21 PST
(In reply to Jan Honza Odvarko from comment #44)
> What is currently the target Firefox version for the patch?
Take a look at the tracking flags on the top right.
Comment 47 Jan Honza Odvarko [:Honza] 2012-01-10 07:10:59 PST
(In reply to Scoobidiver from comment #46)
> (In reply to Jan Honza Odvarko from comment #44)
> > What is currently the target Firefox version for the patch?
> Take a look at the tracking flags on the top right.
Ah, I see, thanks!
Honza
Comment 48 Scoobidiver (away) 2012-01-10 08:00:53 PST
*** Bug 716849 has been marked as a duplicate of this bug. ***
Comment 49 Swarnava Sengupta (:Swarnava) 2012-01-15 21:57:44 PST
One of Fx user reproduce this issue even on Firefox 10 beta 4.
 Crashid here: https://crash-stats.mozilla.com/report/index/1dfcda37-46cb-4585-a164-ae5162120115
Comment 50 Ariel Shkedi 2012-01-16 01:15:04 PST
Created attachment 588827 [details]
A script that always triggers a crash for me

I attached a file that always triggers a crash for me. However, I don't know if it's the same crash this report is dealing with or not.
Comment 51 Swarnava Sengupta (:Swarnava) 2012-01-16 01:26:04 PST
*** Bug 718378 has been marked as a duplicate of this bug. ***
Comment 52 Ariel Shkedi 2012-01-16 02:02:30 PST
I just tried firefox 10 (beta) and it still crashes.
Comment 53 Ariel Shkedi 2012-01-16 02:12:11 PST
(Sorry for not doing this in one message.)

I just tried it in firefox 11 and it crashes there as well. I have adjusted the flags. (I can't seem to find firefox 12 or I would test there as well.)

See: https://crash-stats.mozilla.com/report/index/cfd73cd5-b41e-41cc-a5cd-791bf2120116

(And I hope the crash I am checking is the same as the crash in this bug - but I believe it is.)
Comment 54 Swarnava Sengupta (:Swarnava) 2012-01-16 02:17:50 PST
(In reply to Ariel Shkedi from comment #53)
> (Sorry for not doing this in one message.)
> 
> I just tried it in firefox 11 and it crashes there as well. I have adjusted
> the flags. (I can't seem to find firefox 12 or I would test there as well.)
> 
> See:
> https://crash-stats.mozilla.com/report/index/cfd73cd5-b41e-41cc-a5cd-
> 791bf2120116
> 
> (And I hope the crash I am checking is the same as the crash in this bug -
> but I believe it is.)

Which version of Firebug are you using? Please try with 1.9.0, 1.10a1
Comment 55 Ariel Shkedi 2012-01-16 02:29:28 PST
(In reply to Swarnava Sengupta (:Swarnava) from comment #54)

> Which version of Firebug are you using? Please try with 1.9.0, 1.10a1

I used firebug 1.9.0 for all the tests.

I just tried firebug 1.10a1 and it still crashes. (But even if firebug works around it, that doesn't actually solve the crash.)
Comment 56 Blake Kaplan (:mrbkap) 2012-01-16 02:56:18 PST
Ariel, can you test with a trunk build? I believe your stack should be fixed by bug 712289 which hasn't yet received approval to land on Aurora or Beta.
Comment 57 Ariel Shkedi 2012-01-16 03:21:32 PST
(In reply to Blake Kaplan (:mrbkap) from comment #56)
> Ariel, can you test with a trunk build? I believe your stack should be fixed
> by bug 712289 which hasn't yet received approval to land on Aurora or Beta.

I tried trunk (12) and the test I attached here no longer crashes it, unfortunately the original page I extracted the test from DOES cause a crash.

See: https://crash-stats.mozilla.com/report/index/cd8c097d-279a-456b-b853-624b92120116

Right now I am not able to try to find a small testcase that I can upload, but I will see if I can do so later.
Comment 58 Scoobidiver (away) 2012-01-16 03:33:40 PST
(In reply to Blake Kaplan (:mrbkap) from comment #56)
> Ariel, can you test with a trunk build? I believe your stack should be fixed
> by bug 712289 which hasn't yet received approval to land on Aurora or Beta.
bug 712289 will land in 12.0a1/20120116.
Comment 59 Swarnava Sengupta (:Swarnava) 2012-01-16 04:28:16 PST
Why its not land on 10 beta or 11 aurora? The supprt forum and army of awesome are covered with this issue, firefox 12 means we should wait 4month? :|
Comment 60 Scoobidiver (away) 2012-01-16 04:45:50 PST
(In reply to Swarnava Sengupta (:Swarnava) from comment #59)
> Why its not land on 10 beta or 11 aurora?
The patch of this bug landed in Aurora and Beta.
For bug 712289 which its patch is recent, risks are evaluated by release drivers as there may be unknown side effects (high volume crashes in other places, for instance). For a null check risks are low, for many changed lines risks are high.

> The supprt forum and army of awesome are covered with this issue
It's about Firefox 9.0.1.

It's #63 top browser crasher in 10.0b4, #48 in 11.0a2 and #25 in 12.0a1.
On Mac OS X where it is a top crasher in 9.0.1, it's #19 top crasher in 10.0b4 and #54 in 11.0a2.
Comment 61 Ariel Shkedi 2012-01-17 22:49:05 PST
(In reply to Scoobidiver from comment #58)
> bug 712289 will land in 12.0a1/20120116.

I just tested 12.0a1/20120117 and it did not crash :)

What will it take to get the fix into 10? I was forced to downgrade back to 8 because I could not stand the constant crashes in 9.

Please tag this bug as a release blocker for 10.
Comment 62 Jan Pejša 2012-01-17 23:18:31 PST
(In reply to Ariel Shkedi from comment #61)
> I was forced to downgrade back to 8 because I could not stand the constant crashes in 9.

I'm using same "workaround" (but it is not so easy to get Firefox of version 8). Sometimes I also debug in Google Chrome/Chromium.
Comment 63 Michael Hase 2012-01-18 01:22:56 PST
(In reply to Jan Pejša from comment #62)
> (In reply to Ariel Shkedi from comment #61)
> > I was forced to downgrade back to 8 because I could not stand the constant crashes in 9.
> 
> I'm using same "workaround" (but it is not so easy to get Firefox of version
> 8). Sometimes I also debug in Google Chrome/Chromium.

Same here, reproducable crashes mit firefox 9.0.1 and firebug 1.9.0 when debuggung extjs applications. This is with scientific linux 6.1. Downgraded to firefox 8.0.1 and disabled automatic updates. A download archive for older firefox releases would be helpful.
Comment 64 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-18 01:46:56 PST
(In reply to Michael Hase from comment #63)
> A download archive for older firefox releases would be helpful.

http://ftp.mozilla.org/pub/mozilla.org/firefox/releases/

And for really old stuff:

ftp://archive.mozilla.org/pub/mozilla.org/firefox/releases/

Generally, please don't fill up this bug report with just "Me too, please put this in Firefox 10" comments.

Instead of switching to an older version, consider switching to beta, aurora or development builds: you'll be able to spot issues like this sooner and help us fix them before they end up in the release build! :-)

(eg. http://www.mozilla.org/en-US/firefox/beta/, http://www.mozilla.org/en-US/firefox/aurora/ , http://nightly.mozilla.org/ )
Comment 65 Michael Hase 2012-01-18 02:33:44 PST
(In reply to Gijs Kruitbosch from comment #64)
> (In reply to Michael Hase from comment #63)
> > A download archive for older firefox releases would be helpful.
> 
> http://ftp.mozilla.org/pub/mozilla.org/firefox/releases/

9.0.1 is oldest release, 8.0.1 is empty ;-)

> 
> And for really old stuff:
> 
> ftp://archive.mozilla.org/pub/mozilla.org/firefox/releases/

Great, thanks, this is what I needed.

> 
> Generally, please don't fill up this bug report with just "Me too, please
> put this in Firefox 10" comments.
> 
> Instead of switching to an older version, consider switching to beta, aurora
> or development builds: you'll be able to spot issues like this sooner and
> help us fix them before they end up in the release build! :-)
> 
> (eg. http://www.mozilla.org/en-US/firefox/beta/,
> http://www.mozilla.org/en-US/firefox/aurora/ , http://nightly.mozilla.org/ )

Ok, granted. I'll try to help a little bit, testpage is
http://dev.sencha.com/deploy/ext-4.0.7-gpl/examples/portal/portal.html

Results with different firefox versions and firebug 1.9.0 with active console and javascript debugger:

firefox 8.0.1: runs
firefox 9.0.1: crash
firefox 10.0b2: crash
firefox 11.0a2 (2012-01-17): runs
firefox 12.0a1 (2012-01-17): runs (firebug 1.10a1)

So my crash seems to be fixed in firefox 11 (aurora) and 12 (nightly).
Comment 66 Robert Kaiser 2012-01-18 05:18:30 PST
(In reply to Ariel Shkedi from comment #61)
> What will it take to get the fix into 10?

According to comment #40, it has been taken for 10 (which is in the mozilla-beta repository right now) already, the first .

(In reply to Michael Hase from comment #65)
> firefox 10.0b2: crash

Please try the current beta, which is 10.0b4 (if you downloaded a previous beta, just use the update functionality [Help > About Firefox] and it will be updated to 10.0b4 automatically).


That said, please no "me too" comments, we already know that 9 crashes. 10 is headed for release on January 31, and current betas should have the fix. If you can verify that, you can help us make sure that we'll be good with the release on the 31st.
Comment 67 Ariel Shkedi 2012-01-18 05:30:25 PST
(In reply to Robert Kaiser (:kairo@mozilla.com) from comment #66)

> Please try the current beta, which is 10.0b4

Just tried it, and it still crashes.

https://crash-stats.mozilla.com/report/index/05d701b7-5ed6-4860-ac80-b909c2120118

> 10 is headed for release on January 31, and current betas should have the fix.
> If you can verify that, you can help us make sure that we'll be good with
> the release on the 31st.

But current betas do not have the fix, that's why all the worried comments. This bug needs to block 10 - if needs to be delayed, then delay it, but it can not be released like this.
Comment 68 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-18 05:39:39 PST
(In reply to Ariel Shkedi from comment #67)
> (In reply to Robert Kaiser (:kairo@mozilla.com) from comment #66)
> 
> > Please try the current beta, which is 10.0b4
> 
> Just tried it, and it still crashes.
> 
> https://crash-stats.mozilla.com/report/index/05d701b7-5ed6-4860-ac80-
> b909c2120118
> 
> > 10 is headed for release on January 31, and current betas should have the fix.
> > If you can verify that, you can help us make sure that we'll be good with
> > the release on the 31st.
> 
> But current betas do not have the fix, that's why all the worried comments.
> This bug needs to block 10 - if needs to be delayed, then delay it, but it
> can not be released like this.

The patch isn't in 10b4, it will be in b5, see http://hg.mozilla.org/releases/mozilla-beta/pushloghtml - b5 hasn't been cut yet, it seems, so you'll need to exercise some patience or make your own build off the mozilla-beta repo.

Anyway, even if that page would still crash with the same signature, repeating here that you would like this to block the release of Fx10 without additional information doesn't help. Instead, try to provide a testcase which still crashes with the patch applied, and good steps to reproduce. For now, please wait and check what happens in a Fx10 build *with* the fix.
Comment 69 Robert Kaiser 2012-01-18 06:07:51 PST
(In reply to Gijs Kruitbosch from comment #68)
> The patch isn't in 10b4

How do you determine that? From all I'm seeing, 10b4 should have the fix. It was checked in on the 9th, b4 was being created on the 11th. And rev 4b1ae47f1dd3 is in https://hg.mozilla.org/releases/mozilla-beta/rev/FIREFOX_10_0b4_RELEASE from what I can tell.
But bp-05d701b7-5ed6-4860-ac80-b909c2120118 looks a lot like 10b4. So I'm really confused now.
Comment 70 :Gijs Kruitbosch (away 26-29 incl.) 2012-01-18 06:10:56 PST
(In reply to Robert Kaiser (:kairo@mozilla.com) from comment #69)
> (In reply to Gijs Kruitbosch from comment #68)
> > The patch isn't in 10b4
> 
> How do you determine that? From all I'm seeing, 10b4 should have the fix. It
> was checked in on the 9th, b4 was being created on the 11th. And rev
> 4b1ae47f1dd3 is in
> https://hg.mozilla.org/releases/mozilla-beta/rev/FIREFOX_10_0b4_RELEASE from
> what I can tell.
> But bp-05d701b7-5ed6-4860-ac80-b909c2120118 looks a lot like 10b4. So I'm
> really confused now.

There was an additional patch from bug 712289 (security-sensitive, I can't see it), see comment #56 and comment #58, which got landed later: http://hg.mozilla.org/releases/mozilla-beta/rev/59d8e5608ab7 . Sorry for the confusion I may have caused by just saying 'the' patch...
Comment 71 Robert Kaiser 2012-01-18 09:16:14 PST
(In reply to Gijs Kruitbosch from comment #70)
> There was an additional patch from bug 712289 (security-sensitive, I can't
> see it), see comment #56 and comment #58, which got landed later:
> http://hg.mozilla.org/releases/mozilla-beta/rev/59d8e5608ab7 . Sorry for the
> confusion I may have caused by just saying 'the' patch...

Ah, forgot that. In that case, I hope people here will be testing 10b5, which is being created within the next days and should be out for the public on Friday, I hope.
Comment 72 Michael Hase 2012-01-18 09:22:28 PST
(In reply to Robert Kaiser (:kairo@mozilla.com) from comment #71)
> (In reply to Gijs Kruitbosch from comment #70)
> > There was an additional patch from bug 712289 (security-sensitive, I can't
> > see it), see comment #56 and comment #58, which got landed later:
> > http://hg.mozilla.org/releases/mozilla-beta/rev/59d8e5608ab7 . Sorry for the
> > confusion I may have caused by just saying 'the' patch...
> 
> Ah, forgot that. In that case, I hope people here will be testing 10b5,
> which is being created within the next days and should be out for the public
> on Friday, I hope.

Thanks, I'll test 10b5 as soon as it's out. My tests in comment 65 were really done with 10b4, not 10b2. Just confused the version numbers.
Comment 73 Jan Pejša 2012-01-18 11:46:33 PST
I want to point that many of Firefox users want to use the browser. If using of the browser is not possible, they have strong reason to try (or to use) another browser. (Do you really want this?)

I think that reproducible crash should be fixed ASAP and not in next release (in Firefox 10?).

Motto:
- I do not want to test beta releases of the browser.
- I just want to use it.
- Final version MUST work.

(I apologize for unrelated comment - but I'm not satisfied with current solution)
Comment 74 Robert Kaiser 2012-01-18 12:19:21 PST
(In reply to Jan Pejša from comment #73)
> I think that reproducible crash should be fixed ASAP and not in next release
> (in Firefox 10?).

Firefox 10 is ASAP. We probably wouldn't be able to deliver a well-tested update to many users in a week, and in less than two weeks we are releasing Firefox 10, which has the fix.
Comment 75 Scoobidiver (away) 2012-01-18 14:05:26 PST
(In reply to Jan Pejša from comment #73)
> If using of the browser is not possible, they have strong reason to try (or to 
> use) another browser. 
How many new users will have Firebug installed (correlated at 96%)?
As Firebug is for Web developers, I think they can try 10.0b5 if 9.0.1 is unusable for them.
Comment 76 Ariel Shkedi 2012-01-18 14:32:04 PST
(In reply to Gijs Kruitbosch from comment #68)

> The patch isn't in 10b4, it will be in b5

I just tested 10b5 candidate dated Jan 18 and it did not crash :)

From: http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/10.0b5-candidates/ if anyone else wants to test.

> Instead, try to provide a testcase which still crashes with the patch
> applied, and good steps to reproduce.

I tested both the testcase that I uploaded and http://dev.sencha.com/deploy/ext-4.0.7-gpl/examples/portal/portal.html and was not able to trigger a crash.

So all seems to be well.

Everyone else who is experiencing this: Please test the candidate as well if you can.
Comment 77 Michael Hase 2012-01-19 00:26:42 PST
(In reply to Ariel Shkedi from comment #76)
> I just tested 10b5 candidate dated Jan 18 and it did not crash :)
> 
> From:
> http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/10.0b5-candidates/ if
> anyone else wants to test.
> 
> > Instead, try to provide a testcase which still crashes with the patch
> > applied, and good steps to reproduce.
> 
> I tested both the testcase that I uploaded and
> http://dev.sencha.com/deploy/ext-4.0.7-gpl/examples/portal/portal.html and
> was not able to trigger a crash.
> 
> So all seems to be well.
> 
> Everyone else who is experiencing this: Please test the candidate as well if
> you can.

Tested the sencha examples and my own app which triggered the crash too. With 10.0b5-candidate Jan 18 and active firebug everything works well. Thanks for the great work.
Comment 78 Scoobidiver (away) 2012-01-19 01:21:53 PST
Verified in 10 Beta 5 in comment 76 and comment 77.
Verified in 11 Aurora and 12 in comment 65.
Comment 79 Robert Kaiser 2012-01-19 05:28:41 PST
(In reply to Ariel Shkedi from comment #76)
> I just tested 10b5 candidate dated Jan 18 and it did not crash :)

Yay, thanks. This will go out to beta users on Friday (unless QA finds problems) and to release users on Jan 31, so we are good. Anyone seeing this problem on 9 should either deactivate Firebug until 31st or just go and use Beta until then.
Comment 80 Scoobidiver (away) 2012-01-23 06:26:34 PST
*** Bug 720348 has been marked as a duplicate of this bug. ***
Comment 81 KLEIN Stéphane 2012-01-24 13:08:43 PST
I've the same bug with Firefox 9.0.1 and 1.9.0 on Ubuntu 11.10
Comment 82 Scoobidiver (away) 2012-01-24 14:34:53 PST
(In reply to KLEIN Stéphane from comment #81)
> I've the same bug with Firefox 9.0.1 and 1.9.0 on Ubuntu 11.10
It's fixed in Firefox 10.0b5.

Note You need to log in before you can comment on or make changes to this bug.